The “Out-of-Africa” theory is based on the hypothesis that a small group of Homo sapiens left Africa some 100 000 to 80 000 years ago, the “second exit”, because there was a first exit of Homo erectus much earlier. Those who left Africa in the second exit became ancestors to all non-African populations on this planet.

Human language is specific to Homo sapiens, a relevant aspect to their evolutionary success. The major questions is: did human language emerge just once (the monogenesis theory of human language), or has it emerged independently several times (the polygenesis theory) to have given rise to the more than 6 000 languages that have been spoken on this planet? It could have emerged as many as 200 times, if we base the assumption on the confirmed language families, or much less, if we accept more recently proposed 34 (or just 7) language “super-phyla”, or language could have emerged only once.

In the African situation, there are four macro-families: Khoisan, Niger-Congo, Nilosaharan, and Afroasiatic (which is the only one that has been made a candidate for inclusion in the non-African family called Nostratic). Three out of four of these major linguistic groupings in Africa don’t form part of any major language regroupings, and that could tell us something when we look back into the history of the human language. That’s why I refer to the three Nilosaharan, Niger-Congo and Khoisan languages as “residual” African language macro-families, as opposed to those language families that emerged after modern humans had left the African continent.

With the double anniversary of Charles Darwin last year, and 150 years since the publication of On the Origin of Species, genetics finally meets linguistics – this intersection of the fields of language studies and human genetics is very new. The most fascinating thing about linguistics meeting genetics is that the biological family trees of modern man and linguistics show numerous matches and only a few mismatches. In cases of mismatches, we either have genetic assimilation where the language stayed the same; or genetics have remained the same but the populations shifted to other languages found in their present areas.

One such mismatch is found in Africa, with the Ethiopians. Genetically, they’re classified as African, most closely related to the Khoisan. But linguistically, Ethiopians speak Afroasiatic languages, predominantly spoken by populations that are biologically Caucasian. One explanation is that there’s been an early language shift from African populations in Ethiopia to a linguistic super-stratum, Afroasiatic.

Similar cases are found in northern Europe with the Lapps (genetically Caucasian, linguistically Uralic), with the north-western American Indians (genetically American like the other American Indians who speak Amerindian languages;  linguistically, they speak languages that rather point back into Asia); and in New Guinea, where genetically people are south-east Asian and so should be more closely related to Australians. But linguistically they don’t speak a language that would go with that genetic label, they speak so-called Indo-Pacific languages that are much closer to those of Melanesian populations.

This raises questions that have never been asked before because there was no reason to ask them since we’d never tried to match the biological and linguistic family trees, not before modern ground-breaking research in human genetics.

Remember language shift is different from biology and human genetics, in that it can happen easily within one or two generations. Think of immigrant populations, it usually takes two or three generations to become linguistically part of the host community, whereas genetics takes a much longer time to assimilate.

Natural processes lie behind all language change from generation to generation which would explain the divergence between today’s world languages, even if they had sprung from one common ancestral language. But could we ever hope to prove a single origin of human language? This would depend on the linguistic methodology that we use.

Could all humankind be considered African and speaking African languages?  Indeed, if all humankind are offspring of one small group of emigrants from Africa, of one common ancestress who we call biological Eve some 100 000 to 85 000 years ago; and if language is indeed specific to the species Homo sapiens and therefore originates in Africa; and if it was already fully developed as human language by the time of the second exit of modern humans from Africa – then all humankind could indeed be speaking regional variants of possibly one common “mother tongue”  that originated in Africa.

This assumes that mother tongue was the common language of our ancestress – biological Eve – and her companions at the time of the second exit. This mother tongue would have been an indigenous African language whose dialects, are now spoken by everybody on Earth. But what was its relationship to the residual African languages of Khoisan, Nilosaharan and Niger-Congo?

What we don’t know and wished we knew: how many languages left Africa and how many stayed behind? Was biological Eve part of a linguistically homogenous group who all spoke dialects of mother tongue? What was the relationship of mother tongue to other pre-modern languages in Africa? Or was there a trickling of migrants from Africa with groups speaking different pre-modern languages? If so, which pre-modern African languages were involved? So, did we have a monolingual or multilingual exit? We cannot exclude the possibility that several groups speaking different languages migrated out of Africa because there was already linguistic diversification on the continent before the second exit.

Some of these early language families in Africa may have disappeared with little or no traces left in other languages. We have an interesting case in Africa with the pygmies. There are no pygmy languages today but there’s good reason to assume that at one time they had their own languages. And there may have been other language families that have died out without leaving trace.

Here are two different scenarios, although I don’t know which is correct. The first assumes that the emergence of language predates or coincides with the emergence of Homo sapiens. “Proto-Human” would be the ancestral form of language of Homo sapiens, which diversified over time, giving rise to “proto-mother tongue”. So proto-mother tongue would then have been one of several languages that developed over time. This would mean that before the second exit Africa was home to several language families, one of them mother tongue, probably amongst people inhabiting the shores of modern Eritrea, Djibouti, Somalia, from where they eventually migrated across the Red Sea (at the time there was an ice age, a lot of water was tied up in ice caps so the sea had fallen dry).  Eve’s group, which spoke mother tongue, took their language out of Africa, likely without leaving linguistic traces behind. Then all the other languages of the world developed out of this.

This assumes that one human language diversified into several proto-languages with some becoming extinct, while others however survived. Proto-Khoisan, proto-Nilosaharan, proto-Niger-Congo would have evolved to become present day Khoisan, Nilosaharan, Niger-Congo languages in Africa, forming the residual African languages. One of the daughter languages was proto-mother tongue which was spoken by those who took part in the second exit from Africa. It was taken out of Africa and became the ancestral language to all the “out of Africa” languages.

The alternative is that mother tongue was the ancestral human language, rather than the daughter of it, and that it diversified before the second exit. Some of these early language families became extinct, others survived. Under this scenario, all or many of the speakers of early language families had access to the highlands of Ethiopia and the shores of modern Eritrea, Djibouti, Somalia, from where some of them eventually migrated across to the Arabian Peninsula, while other speakers of the same language families remained behind in Africa. Eve’s group would have been part of a sequence of many different linguistics groups who migrated out of Africa and took their languages along, which later developed outside Africa into the language families that we know today.

For Africa, it is interesting to note that in addition to the residual African language macro-families Khoisan, Nilosaharan, and Niger-Congo, there are the Afroasiatic languages in North-Eastern Africa. These could be assumed to have originated in the Fertile Crescent outside Africa where the invention of agriculturalism and pastoralism (“Neolithic Revolution”) took place some 10 000 years ago. Non-linguistic evidence points towards a hypothesis according to which the technologies of food production arrived in Africa together with the speakers of Afroasiatic languages and, spreading slowly across the then green Sahara, another 7 000 years later sparked off the so-called Bantu Expansion from West Africa over much of Eastern and Southern Africa.

Prof Ekkehard Wolff is the retired Chair of African Languages and Linguistics at the University of Leipzig, Germany. He delivered the lecture Human journey out of Africa – a perspective from language studies on 11 March 2010 as part of the 2010 Darwin Seminars hosted by the African Genome Education Institute.